ossec-agentd consumes 100% CPU - requires restart

[kapshure]

hi, we are on 3.6 (server + agents) and recently within the past few months, have seen a rampant increase in the number of agents that start experiencing 100% CPU load due to ossec-agentd.

We do have a lot of disconnected agents, not active, and many that have never connected. Been seeing lots of "error 1403, error 1213, error 1407".. currently we're just restarting the ossec-agentd process, but is there any idea from the dev team what could be causing this?

Is the ossec-agentd single-threaded, or is it multi-threaded? the agents are mostly on CentOS 6-8 x64 any help is appreciated